The physical and chemical environment of a floodplain needs to be assessed to define conservation targets for restoring it to species-rich meadows from agricultural land. A straightforward technique, widely applicable by site managers for assessing the suitability of the hydrological and hydro-chemical regime of a floodplain for wet grassland restoration, has been tested by examining the feasibility of restoring plants characteristic of NVC MG4 and MG8 communities to the Castle Meadows, Wallingford (Oxfordshire, UK). Hydro-chemical suitability has been assessed by comparing phosphorus concentrations with species-rich meadows nearby. The flooding regime was estimated based on a rating curve and a digital elevation model and groundwater levels were measured monthly in dipwells and piezometers. The hydrological regime was then compared with published reference guidelines for communities of conservation interest. For the Castle Meadows, the maximum duration of flood events in autumn and winter exceeded MG4 and MG8 species requirements across half of the site, while the depth of the groundwater table in summer exceeded species requirements in the other half. It was shown that, depending on topography, MG5 or MG13 may be more realistic vegetation targets.

Abstract. We developed a high-resolution MIKE SHE/MIKE 11 model of a 231.3 ha headwater catchment in the granitic uplands of the French Massif Central to estimate the contribution of groundwater upwelling to the water balance of the Dauges mire, an acidic valley mire of international importance for nature conservation. We estimated that groundwater upwelling from the underlying weathered granite formations – mostly an approximately 55 m deep fissured zone – provides 27.1 % of total long-term inflows to the mire. This contribution increases to 37.2 % in September when total inflows are small. Overland boundary inflow accounts for an average of 40.2 % of total inflows. However, most of this originates from groundwater seepage through mineral soils along the mire margins or in small non-channelised valleys upslope of the mire. A sensitivity analysis showed that model performance in terms of the simulation of mire groundwater levels was most sensitive to parameters describing the mineral soils and weathered granite formations rather than the overlying peat layer. Variation partitioning demonstrated that groundwater upwelling was the most important factor driving simulated monthly groundwater table depth within the mire. Sustained groundwater upwelling maintains the mire water table close to or at ground level for most of the year. As a result, precipitation and overland boundary inflows quickly leave the wetland as saturation-excess runoff. There was close agreement between the observed distribution of mire habitats and areas where the simulated long-term groundwater seepage rate was larger than zero in September. Our results demonstrate that, contrary to the assumed small contribution of groundwater to the hydrology of hard-rock regions, groundwater upwelling from underlying weathered formations can be a quantitatively important and functionally critical element of the water balance of valley mires in granitic headwater catchments. These results have important legal and management implications.

Abstract. The contribution of groundwater to the hydrology of hard rock regions has long been assumed to be small. This is being progressively challenged and conceptual hydrological models of headwater wetlands in these regions may need to be revised. We developed a high-resolution MIKE SHE/MIKE 11 model of a 231.3 ha headwater catchment in the granitic uplands of the French Massif Central to estimate the contribution of groundwater upwelling to the water balance of the Dauges mire, an acidic valley mire of international importance for nature conservation. We estimated that groundwater upwelling from the underlying granite weathering formations – mostly an approximately 55  m deep fissured zone – provides 27.1 % of total long-term inflows to the mire. This contribution increases to 37.2 % in September when total inflows are small. Overland boundary inflow accounts for an average of 40.2 % of total inflows. However most of this originates from groundwater seepage through mineral soils along the mire margins or in small unchannelized valleys upslope of the mire. A sensitivity analysis showed that model performance in terms of the simulation of mire groundwater levels was most sensitive to parameters describing the mineral soils and granite weathered formations rather than the overlying peat layer. Variation partitioning showed that groundwater upwelling was the most important factor driving simulated monthly groundwater table depth within the mire. Sustained groundwater upwelling maintains the mire water table close to or at ground level for most of the year. As a result, precipitation and overland boundary inflows are mostly evacuated as saturation-excess runoff. There was close agreement between the observed distribution of mire habitats and areas where the simulated long-term groundwater seepage rate was larger than zero in September. Groundwater upwelling from the underlying weathered formations can be a quantitatively important and functionally critical element of the water balance of valley mires in granitic headwater catchments. These results have important legal and management implications.

We thank Anonymous Referee #1 for their comments and for suggestions on the original paper. We believe that the paper will be improved by responding to the issues that they raise and by incorporating their suggested revisions.

No abstract